What the new 5DII is telling us

I'm using Nikon primes on my 5D!
Must be. With the D700 you're locked into a bunch of screw-driven
legacy lenses and significantly less sensor resolution. Heh...
I can understand that you Canonites are envious. You do not have
access to all those legacy lenses that we can buy rather cheaply. But
do not despair, Canon makes many good lenses and now you can buy
Zeiss, you are almost there.......
--
life is raw
--
Galleries and website: http://www.whisperingcat.co.uk/mainindex.htm
 
We haven't seen the noise performance of the 5dII's sensor yet. I seriously doubt Canon would release a sensor that did not perform at least as well as the 5D's excellent sensor.
--
Phil Flash
SF, CA USA

It's not the camera. It's you.

Stuff I own in my profile.
 
people who buy a D40 don't care / aren't supposed to care about primes, full frame or brand loyalty. They just made it to high shool ;-)
--
Buying a Nikon doesn't make you a photographer, it makes you a Nikon owner.
 
Canon profoundly leads Nikon in sensor development.
Really? The difference in IQ between the D3 and 1DsIII is just the pixel count. Let's wait and see what the high ISO is for this new sensor. At most it'll match the D3/D700's, with pixel count advantage. Is that "profoundly" leading? Talk about hyperbole.
Nikon profoundly leads Canon in designing camera bodies with useful
features and making them available to more photographers besides
those able to pony up for so-called flagship models.
Really? The difference between the D3 and the 1DIII is about nothing, and wait for the new 1DIV. The D700 is a small PJ/sports camera, the 5DII is more of an advanced enthusiast camera, different public. I expect the 5DII to be selling for about 2,00USD in 6 months. Nikon will likely have a FF under 2,000 in at most 2 years. Again, Canon can match any body specs Nikon has if they want, at any moment, and likely at a cheaper price.
5dII's body is almost unchanged from 5dmkI. That would be like Nikon
upgrading the sensor in the D70 to the current full frame sensor, but
not changing any of its features. Silly.
Not silly, it'll sell for about 2,000 soon and many enthusiasts and budding pros will get it.
I hate Canon's philosophy of protecting high end bodies by stripping
mid-level bodies of features and hobbling them with mediocre AF and
metering and ergos.
Read you own line below, is it the camera?
--
Phil Flash
SF, CA USA

It's not the camera. It's you.

Stuff I own in my profile.
--
Regards, Renato.
http://www.flickr.com/photos/11435304@N04
OnExposure member
http://www.onexposure.net/

You cannot depend on your eyes when your imagination is out of focus
(Mark Twain)
 
IMO it'll be selling for 2,000USD in six months, and will have strong appeal to the advanced enthusiast and some budding pros.

It's not an action camera as the D700 is, this one a cheaper/lighter version of the D3, but a simpler camera with excellent image quality.

I'm waiting to see if I'll go with FF, because I shoot mostly in the WA range (see my pics in link below), and this could be my next camera, as well as the D700. Likely I'll go with the D700 or wait for a D90 with FF sensor, likely to be here in a couple of years.

For pros, the D700 certainly looks more attractive, unless you're a landscape photog. For weddings/events, the speed, AF, and likely still, better high ISO, are worth more than more megapixels.

--
Regards, Renato.
http://www.flickr.com/photos/11435304@N04
OnExposure member
http://www.onexposure.net/

You cannot depend on your eyes when your imagination is out of focus
(Mark Twain)
 
Actually, there are already a few full size samples out there. ISO25,600 is awful, as expected. ISO6400 looks fine to me.
We haven't seen the noise performance of the 5dII's sensor yet. I
seriously doubt Canon would release a sensor that did not perform at
least as well as the 5D's excellent sensor.
 
Exactly..this is just gimmicky nonsense....still and moving images are separate art forms.

If I ever wanted a video camera I would buy one..but I have been taking still pictures for 44 years since I was 8 and have never felt the need!!!

In fact I am put of a still camera that offers "movies"!

I have little doubt though that the marketing men will be convincing people this is the next must have and I wonder how long it will be before my wedding customers are asking if they can have "just a few minutes of video please" because .."well, it's sumthink difrint innit!"
 
Canon profoundly leads Nikon in sensor development.
Not really. The companies are taking two different paths in sensor development. And, by the way, the 5DII sensor is the same as the 1DsIII sensor, so even if you were right on this assertion you're quite late in making it ; ). The only change in the 5DII is to the Bayer layers to make them more transmissive. That kind of change actually worries me a bit, as it means having to retweak converters because the spectral response likely changed.
5dII's body is almost unchanged from 5dmkI. That would be like Nikon
upgrading the sensor in the D70 to the current full frame sensor, but
not changing any of its features.
No, it would not be that equivalent. It would be the equivalent of upgrading the D80 sensor to the current DX sensor, but not really changing any of the camera body features. Oops, that's a D90.

Both Canon and Nikon appear to believe that there are "feature sets" that balance costs (read: profits) versus usability. Nikon has been pretty consistent about having at least one body in the N80 space for some time; Canon appears to be doing the same in the 50D/5D space.
I hate Canon's philosophy of protecting high end bodies by stripping
mid-level bodies of features and hobbling them with mediocre AF and
metering and ergos.
You mean like the D90? Yes, the D200/D300/D700 are pleasantly above such compromises, but Nikon has been doing similar things with the midrange products for some time.

--
Thom Hogan
author, Complete Guides to Nikon bodies (18 and counting)
http://www.bythom.com
 
I hate Canon's philosophy of protecting high end bodies by stripping
mid-level bodies of features and hobbling them with mediocre AF and
metering and ergos.
You mean like the D90? Yes, the D200/D300/D700 are pleasantly above
such compromises, but Nikon has been doing similar things with the
midrange products for some time.
The AF module is probably fairly expensive so putting in a lower end version is probably required to meet the lower price but why does Nikon have to mess up the metering? You've documented how the D80/D40 etc. pay more attention to what is under the selected AF point. Many think it meters "hotter" blowing out highlights. It would take one firmware guy 2 minutes to add an option to "use D200 pro style metering" and cost Nikon nothing except maybe some ROM space which they probably have extra anyway.
 
a double-digit number which the dumb consumer uses to justify buying expensive equipment to photograph his dog and pixel-peep at how sharp his fur is.
 
One of the biggest dollar upgrades for wedding photographers is to add video, and a number of the high end shooters are creating mixed still/video DVDs. (Sorry, OldMaster, but that started happening several years ago.) Putting video in the same body as the stills lets the mid-tier, mid-priced wedding shooter mix it up - in other words, you won't have to buy a $20,000 wedding package to get video, it'll be in reach of the $5-10K wedding package.

Most of the major newspapers (wonder when we'll drop the "papers" part of that word) now ask their photographers to bring back video clips also, particularly for in-depth news stories like documentaries of burn units or new school programs. Most of the content managed by the major newspapers like NY Times now has video add-ons. It's kind of funny watching the photographers juggle their video cams and their still cams, and at least to me (with film school in my background) the results prove that being a good photographer doesn't mean you have a clue about video...

But it's not at all gimmicky nonsense, it's Nikon and Canon giving two of their largest professional markets exactly what they've wished for.

It'll be interesting watching photographers trying to decide if they should be shooting video or stills, yet another way to miss the global cover page image.

Next stop: Red. Why shoot second rate video with a still camera when you can shoot video at 5mp native resolutions and just grab your stills from there?

http://www.red.com/

There are tests going on right now with traditional still image agencies equipping their shooters with Red cameras.
Exactly..this is just gimmicky nonsense....still and moving images
are separate art forms.

If I ever wanted a video camera I would buy one..but I have been
taking still pictures for 44 years since I was 8 and have never felt
the need!!!

In fact I am put of a still camera that offers "movies"!

I have little doubt though that the marketing men will be convincing
people this is the next must have and I wonder how long it will be
before my wedding customers are asking if they can have "just a few
minutes of video please" because .."well, it's sumthink difrint
innit!"
 
What the 5DII and A900 are telling us is that good cameras don't need to be as heavy as the D700!
--
Jal
 
Canon profoundly leads Nikon in sensor development.
Not really. The companies are taking two different paths in sensor
development. And, by the way, the 5DII sensor is the same as the
1DsIII sensor, so even if you were right on this assertion you're
quite late in making it ; ). The only change in the 5DII is to the
Bayer layers to make them more transmissive. That kind of change
actually worries me a bit, as it means having to retweak converters
because the spectral response likely changed.
Colour separation is not the same (1Ds MkIII already started to suffer this problem already). Now it came to the point where we need colour interpolation (something more or less similar to what Foveon-type sensors need), not only debayerization.

--
http://www.libraw.org/
 
Did you fall off your Canon pogo stick, hit your head, and post this before the MRI?

Joe
"On paper, the 5DII competes very well with the D700"

You Are kidding, right?
Must be. With the D700 you're locked into a bunch of screw-driven
legacy lenses and significantly less sensor resolution. Heh...
 
Right now it's telling us the big 3 are very split in what that
market point is.
That's an understatement. From a design standpoint you have to have a target customer and use well in mind. What we've seen from the beginning of digital but don't always discuss is that each of the camera makers has had a somewhat different take on what is important and who will use their product. As I've written elsewhere, I think this is going to get worse, as some companies are mistakenly trying to use technology to artificially create new segments. The problem is that USE always drives segments; technology is only a catalyst.
1) Canon is saying it's lots of pixels in a cheap body with LV and
Video, but not much meat to the feature set.
Canon's philosophy from day one has been pixels, and to a lesser degree image quality. Even at the high end (original 1Ds, for example), it took a long time to get to lithium batteries and other things that I think a truly pro camera required.
2) Nikon is saying it's few pixels, LV and a real solid body, and a
feature set of things that are needed (flash, GPS, tank build)
Nikon's philosophy has been improvement the photography-related technology (AF, metering, flash, etc.). Despite the D3/D300/D700, I'm still not 100% convinced that Nikon fully understands or pursues "ultimate image quality." Their choices for the AA filter, default sharpening, default color, and a host of other pieces are not optimal, and have NEVER been.
3) Sony is saying it's high pixels in a traditional body (huge View
finder, no LV, no Video), but giving you stabilization to work with
Primes that target users can't get stabilized in other brands
(portrait, landscape type lenses).
Sony's philosophy is hard to pin down. The cameras have been all over the place, IMHO. I suspect it is a resource problem, not a targeting problem. They're trying to do a lot in a short amount of time with I think fewer design/engineer resources, and when that happens things get left out and easy optimizations take precedence over difficult ones.
4) Pentax is saying "damn we need to do something"
The Hoya takeover appears to have done what I thought it would: stall a generation of product. Unfortunately, that's not a good thing when the market is moving so rapidly.
If your going to do video, get a real video camera.
I'm going to disagree with you. But I'm going to agree that the choices so far seem really problematic. Let's start with Nikon: adding video to a truly consumer camera seems likely to stick. The problem is that Nikon's implementation isn't good enough. The consumer wants "automatic with overrides." The D90 is automatic with overrides on the still end, but "mostly manual" on the video end. There's a dissonance between the level of the video and still features.

The 5DII is a little better. Here we have essentially the highest possible image quality for stills at the moment, coupled with the highest possible image quality for video (yes, I know that better video is possible, but in terms of what we normally mean by "video," 1080P is the highest ubiquitous format). What strikes me is the compromises in the camera (shutter, AF, auto gain for sound, etc.). There's a little too much "pushing pixels" that now doesn't seem to match up with the rest of the capabilities. It's a little too much like a 4 liter V8 in a compact car.
I have to wonder
how well the Canon Video camera that takes EOS lenses has sold over
the years.
Quite well, actually. The difference here is like the D90 (and the Red): the larger sensor provides a different (and to many preferred) "look" than the small sensored video cameras we've had to date.
The Sony is also bound to get in a price war with Canon at some point
too.
Actually, Canon is the instigator of most of the digital price wars. They've been consistently taking lower product margins on everything except the 1D products in order to press volume and hold market share. The pricing of the 5DII looks no different, as it undercuts Sony just as the Sony comes to market (and has features the Sony doesn't have).
And I'm pretty sure Sony will win on getting the price lowest.
Not sure why you say this. I'd say the opposite. First, ask yourself why the 5DII uses many of the same parts as the 5D? When you play price pressure games parts volume starts to come into play. Canon has more ability to spread this across multiple products than Sony right now.
Canon had the opportunity to show their cards last, so they went with
a lower price.
Actually, Canon was second. Sony made a big (and classic) mistake by pre-announcing the A900, including price point. That gave everyone else a clear target. We've yet to hear from Nikon, Pentax/Samsung, and possibly one other FX producer, but the target is clear: match or beat 20mp, video, capable camera at US$2699 or better.
Though it's not like 40D/50D has lowered D300 pricing,
True, but I've always said that the customer target was different for those products.
and the 40D and 50D never pulled the A700 price down either.
The A700 never really had any substantial volume. I have no idea why Sony decided to keep volume low and price high. This says to me that they considered it more a placeholder than one of their more important targets.

--
Thom Hogan
author, Complete Guides to Nikon bodies (18 and counting)
http://www.bythom.com
 
Canon profoundly leads Nikon in sensor development.
Not really. The companies are taking two different paths in sensor
development. And, by the way, the 5DII sensor is the same as the
1DsIII sensor, so even if you were right on this assertion you're
quite late in making it ; ). The only change in the 5DII is to the
Bayer layers to make them more transmissive. That kind of change
actually worries me a bit, as it means having to retweak converters
because the spectral response likely changed.
Colour separation is not the same (1Ds MkIII already started to
suffer this problem already).
Could you elaborate on the problems in the 1Ds MkIII sensor related to colour separation? (or am I mixing up concepts, here)
Now it came to the point where we need
colour interpolation (something more or less similar to what
Foveon-type sensors need), not only debayerization.
I understood that Thom meant that the RGB dye filters in the 5DMkII sensor are less selective (than in the 1DsMkIII version) regarding different wavelengths. That is to say that when earlier sensors and their filters have been designed to maximise selectivity (e.g. red filter cuts out wavelengths below 640 nm as effectively as possible) and thus colour accuracy (I'm speculating out loud), the new 5DMkII colour filters increase the sensor's sensitivity by letting in more photons from other than their respective wavelengths.

Did I understand it correctly? Then, the million dollar question is, does it affect colour accuracy?

You mentioned colour interpolation. While I trust Canon engineers know what they're doing, I fear that there is a price to pay. Are they simply going after sexier ISO ratings or are they really pushing Bayer sensor technology further?

--
regards
Janne Mankila, Finland

14.9.2008
I have three and a half idols.
One, Sirius Black.
Two, Sam Tyler.
The rest are Hugh Lauries.
Yes, it's almost an essay:
http://jannemankila.googlepages.com
 
I understood that Thom meant that the RGB dye filters in the 5DMkII
sensor are less selective (than in the 1DsMkIII version) regarding
different wavelengths.
Colour filters need certain density to have an acceptably sharp cut-off for the bands that are undesirable. Making filters more transparent means the slopes of the transmittance curve are less pronounced, and more of the undesired wavelengths are allowed. In a sense that means that colour response is more muted.
does it affect colour accuracy?
With traditional interpolation, yes, it does. Good algorithms for colour restoration for this application seem to be missing.
Are
they simply going after sexier ISO ratings or are they really pushing
Bayer sensor technology further?
Both, but for now most of what I see is pushing ISO up on smaller pixels with acceptable noise while sacrificing other factors.

--
http://www.libraw.org/
 
I understood that Thom meant that the RGB dye filters in the 5DMkII
sensor are less selective (than in the 1DsMkIII version) regarding
different wavelengths.
Colour filters need certain density to have an acceptably sharp
cut-off for the bands that are undesirable. Making filters more
transparent means the slopes of the transmittance curve are less
pronounced, and more of the undesired wavelengths are allowed. In a
sense that means that colour response is more muted.
An informative summary, by the way.
does it affect colour accuracy?
With traditional interpolation, yes, it does. Good algorithms for
colour restoration for this application seem to be missing.
Surely it's beneficial for Canon to release some kind of spectral charts or adequate interpolation routines to help third party raw converter writers? Having said that, Canon must be highly secretive of any kind of information on the specifics...
Are
they simply going after sexier ISO ratings or are they really pushing
Bayer sensor technology further?
Both, but for now most of what I see is pushing ISO up on smaller
pixels with acceptable noise while sacrificing other factors.
If I understood you correctly, I think that's something that's been happening for a while in the digital P&S cameras. I, for one, am not happy with the trend.

--
regards
Janne Mankila, Finland

14.9.2008
I have three and a half idols.
One, Sirius Black.
Two, Sam Tyler.
The rest are Hugh Lauries.
Yes, it's almost an essay:
http://jannemankila.googlepages.com
 

Keyboard shortcuts

Back
Top